Fardriver ND72680B Encoder Help

[samsonx]

Hello, new to the space but seems like the right place...

I purchased a ND72680B from ebay seller, I requested sin/cos encoder support and this was the version I was advised to get. Unfortunately the documentation I can find (ebay seller pointed me to facebook group) all shows the wiring of HALL sensor wiring and not how to interface a sin/cos encoder.. My motor has a RMB29AC01SS1 – Analogue sine/cosine encoder.

I've put together a wiring table I've seen two different version of the pinout of the fardriver controllers. Here is what I've come up with:


However I get Encoder Hall error single tone repeats every ~1.5s

Any advice would be greatly appreciated!

Thanks!
Sam

 

[absence209]

Hey Samsonx!
I don't have anything pieces of advice or anything but I am in the same boat as you. I was wondering if you got your setup working or if you found any other solutions? Thanks.

 

[samsonx]

Hello @absense209 I was not able to get the sin/cos encoder to interface with the ND72680B, however am in the process of converting the encoder to incremental (TLE5012E1000).

 

[samsonx]

After failing to get sin/cos encoder working, I've built a replacement using the TLE5012B1000E which is an incremental 4096 resolution encoder; which the ND72680B *should* support.

 

[absence209]

Hey Samsonx!

Did this setup end up working?

I just replaced my RMB29 Sin/Cos with a RMB29 Incremental and the motor just wiggles in place.

 

[samsonx]

it does not work! controller throws an error and nothing else. the only setting that appears to even monitor the motor inputs (hall a, b, c) is the setting 0-120HALL, the default setting '1-encoder' (or any other setting) does not. I ordered the part to build a another version of this tle5012 board but in hall mode and not incremental.

your controller also the B variant (eg encoder) right?

 

[absence209]

sorry, i said in my other reply that i was using an incremental sensor. i meant to say commutation. my new sensor outputs a commutation signal, not incremental.


but to answer your question, no, my controller is hall only. which is why I got the RMB29Eb board (2x commutation signal, which emulates a 4 pole configuration).

it doesn't work.

I even went ahead and used an arduino to interpret the original sensors sine cosine signals and use an arctan function to extract the exact angle to recreate a UVW commutation sequence.

nothing worked so far for me.


im actually starting to suspect my motor might not even be a 4 pole, as the only website that states it as 4 pole is thunderstruckev (to be fair theyre the only website that has ANY information on the ME1401 motor at all).


but please keep me updated with everything your doing. I feel like your the only other person who's in the same situation as me right now! best of luck with your project.

 

[amberwolf]

How many magnets does it have? Usually that gets divided by two to get the poles (pole pairs, really).

 

[absence209]

it has 4 poles 8 magnets. its axial flux.

These are the specs:

Home :: Motors DC and PMAC :: ME1401 Splined shaft

This motor has been used in smaller golf cart sized vehicles at 72v rpm needs to be limited to 5000 Tested to 450 amps Peak current Fits many standard 19 tooth splined golf cart transaxles Get it as a kit with our Curtis controller so that we can handle the parameter settings for you 4700 rpm @...

www.thunderstruck-ev.com

 

[absence209]

is it possible the motor is 4 pole pairs? so totaling 8 poles?

that would make sense given a 4 pole (2 pole pair) commutation sequence didn't work.

 

[amberwolf]

You said it yourself : 4 poles, 8 magnets. That's 4 pole pairs, since each magnet should be flipped opposite polarity from the next one over.

Then you just have to figure out if the controller software asks for pole pairs, or just poles... some do one, some do the other, but most just say poles and you have to guess or try different things to figure it out.

 

[absence209]

im so dumb.

just one word costed me 50 bucks. ill purchase an 8 pole sensor and try messing around in the settings of the controller.

Thank you amberwolf!

 

[amberwolf]

Can the controller run sensorless?

(many FOC can)

 

[samsonx]

Can the controller run sensorless?

(many FOC can)

I don't think this unit can.. motorc/Bluetooth app only reports two phase currents and for motor angle selection there is not a disable, just various hall/encoder settings.

it has 4 poles 8 magnets. its axial flux.

These are the specs:

Home :: Motors DC and PMAC :: ME1401 Splined shaft

This motor has been used in smaller golf cart sized vehicles at 72v rpm needs to be limited to 5000 Tested to 450 amps Peak current Fits many standard 19 tooth splined golf cart transaxles Get it as a kit with our Curtis controller so that we can handle the parameter settings for you 4700 rpm @...

www.thunderstruck-ev.com

omg I have this exact same motor! I just today got it "jiggling" as well in hall mode .. I was just reading the datasheet and the tle5012 variant I am using is 5 pole pair as default setting .. I am going to try and connect the data lines and reconfigure as 4 poles. The Motorc app also has a setting for pole pair count.. however from what I've gathered in the fardriver "manual" is that the autolearn /self-learning function should automatically correct the pole pair value. But the hall sensor itself though also needs to match?

Very annoyed that the incremental encoder does not appear to work.. wondering if it was just loaded with a .hex that had encoder as the default option.. or really only supports hall effect mode. or maybe only supports fardrivers official encoder ..

 

[amberwolf]

The ABC/UVW hall sensor type setup normally detects the actual rotor magnets.

For a separate encoder that sends these signals from it's own magnet set instead, it not only must have the same number of poles as the actual rotor does***, it must also be positioned so that the transitions from pole to pole occur at the same points as the actual rotor. If they don't line up, the timing from the halls will be different than that of the phase/rotor magnet, and the drive signals will not be timed for the magnet position of the rotor vs stator teeth.

If the controller can autodetect this offset and correct for it, then that isn't an issue, but if it cannot, then you would have to manually adjust either the angle number (retard/advance) in the controller setup (if available), or else physically rotate the encoder that's connected to the rotor so that it's magnets align with the ones in the rotor.


***or it would need some mulitple of poles as the rotor, and the controller or encoder would have to have a way to adjust for this multiple.

 

[samsonx]

For a separate encoder that sends these signals from it's own magnet set instead, it not only must have the same number of poles as the actual rotor does***, it must also be positioned so that the transitions from pole to pole occur at the same points as the actual rotor. If they don't line up, the timing from the halls will be different than that of the phase/rotor magnet, and the drive signals will not be timed for the magnet position of the rotor vs stator teeth.

If the controller can autodetect this offset and correct for it, then that isn't an issue, but if it cannot, then you would have to manually adjust either the angle number (retard/advance) in the controller setup (if available), or else physically rotate the encoder that's connected to the rotor so that it's magnets align with the ones in the rotor.

@amberwolf thank you for confirmation on the pole pair setting. There is a phase offset setting which I've read will also be updated from a successful autolearn.

I am having a hell of a time configuring the TLE5012B, but also I realize that there is no way to permanently store the pole count setting if I do change it successfully... May need to look at yet another encoder option as I dont want to reconfigure the encoder every single boot..

From the motorc app:


@absence209 note the, "HALL" status item w/ binary word of the encoder state. If you rotate the motor by hand you see these cycle as well?

 

[absence209]

@samsonx, I actually havent used the computer software for fardriver yet, so I haven't seen that setting.

I mainly use the phone app for configuring,

But i do see the hall wires change on my oscilloscope.

 

[samsonx]

I can confirm that the TLE5012B in hall mode set to 4 pole pairs DOES work with the ND72680B! @amberwolf looks like phase offset was determined automatically.

Well, unfortunately running the controller in Enocoder mode (sin/cos or incremental) has been fruitless.. I guess the question is does the 72680B controller I have actually support Encoders? It doesn't appear so.. unless there is configuration or wiring config that I haven't found yet.